Monday, September 08, 2008

The King's Camelopard, or ...

The Royal Nonesuch
Guaranteed to spot
MEDIA BIAS
at the merest CLICK of the MOUSE!
LADIES AND CHILDREN NOT ADMITTED

... And if that don't fetch them, I don't know Arkansas. Which is more or less, I think, what to make of this lede:

In a development that could titillate political partisans of all stripes, a new Web application promising to spot bias in news stories will launch on Monday, Sept. 8, just as this ferociously contested election season shifts into high gear.

Don't get your hopes up too fast; as of a few minutes ago, the new launch date seems to be Sept. 9. But a little slippage in your release date isn't the real problem. The problem is the particular flavor of magic beans that Business Week is trying to sell here. It's a Web app that'll scan all your news for bias, or "spin," thus presumably allowing you to be a better citizen.

What's it going to do? Let's let Business Week tell us!

When turned on in a user's Web browser's toolbar, Spinoculars scans Web pages and spots certain potential indicators of bias. The toolbar also will allow its users to flag phrases in news stories and opine on those called out by other Spinspotter users. The application's algorithms work off six key tenets of spin and bias, which the company derived from both the guidelines of the Society of Professional Journalists' Code Of Ethics and input from an advisory board composed of journalism luminaries.

What sort of tenets might those mighty algorithms use? Let's have a look (interrupted here and there by your editor):

The tenets are: reporter's voice (adjectives used by a journalist that go beyond the supporting evidence in the article);
This is a reliability/validity question: Does the measure we're using provide a consistent index of something, and does that index have some relation to the thing we want it to represent? We have some problems on either count. As a rule, stories that describe "brutal" murders don't provide any evidence to distinguish the murder at hand from any other murder; as Raymond Chandler suggested some decades ago, all murders are brutal. Does that mean those stories are spun, and if so, is an anti-murder spin a bad thing?

That's not to say adjectives -- "fresh-faced maverick Sarah Palin gave another hard-hitting speech last night" -- can't or don't indicate spin. It is to suggest that they're an unreliable indicator of a small part of the spin spectrum. If you want to worry about "reporter's voice," try this: "Smith, who fought in the War on Terror during the invasion of Iraq in 2003, said he would vote for McCain." Not an adjective in the lot, but it's the reporter's voice that makes the invasion part of the War on Terror. If your "voice" algorithm is off hunting adjectives - well, good luck with that supporting evidence, algorithm!

passive voice (example: a story says "bombs land" without stating which party is responsible for them);
Another writer who wouldn't recognize verb voice if it jumped up and he was bitten by it. "Bombs land" is in the "active" voice. Our writer (and presumably our algorithm) is concerned with the absence of agency, which is a legitimate if separate concern: "Bombs land" and "fire breaks out" are both some form of natural disaster. The active voice, alas, is not only good at hiding agency, it's a good way of introducing another set of biases -- particularly for those journalists who want to blame an attack on the Usual Suspects, rather than admitting that nobody knows who carried it out. Beyond that, verb voice leaves a whole raft of "spin" and bias unmeasured. Referring to Iraq as "the war-torn country" leaves open the question of how it got that way, but it doesn't go beyond the "evidence" and it doesn't involve verb voice. What is it we're measuring again?

a biased source (a quoted source's partisanship is not clearly identified);
Worthwhile goal, but we risk confusing facticity with accuracy here. You can pile on the source's name, age, hometown and party affiliation all you want, but that won't turn a lie into something true. Some partisans say things that are true. Some -- I'm thinking of Bill Safire, for one, but you probably have your own examples -- almost invariably don't. If you let a source lie, it doesn't really matter how thoroughly you document his/her "partisanship."

disregarded context (a political rally's attendance is reported to be "massive," but would it have been so huge had the surviving members of the Beatles not played?);
Sorry, missed that gig.

and lack of balance (a news story on a controversial topic gives much more credence to one side's claims).
This used to be known as the "green cheese fallacy": the idea that if your candidate says the moon is an airless rock and the other candidate says the moon is made of green cheese, the hed is going to be CANDIDATES SPAR OVER NATURE OF MOON. And the blunt reality is that on many "controversial" topics, one side's claims ought to be given much more credence because they're ... you know, true.

Smoking kills people and causes cancer. "Intelligent design" is a religious belief, rather than something you teach in a science classroom. Iraq was not an existential threat to U.S. national security when George Bush decided to invade it. These aren't topics that should be given equal weight as legitimate controversies because they aren't controversies. If you pretend there's a controvery when there isn't, that's "spin." Or "bias." Or "outright lying."

Business Week falls for another fallacy when it notes that, although the SpinSpotter founder is a conservative, the CEO is "progressive"! So we can't be biased, right? You've probably seen your local editor make a similar claim: Both sides are mad, so we must be doing something right! Which overlooks, alas, the high probability that having different partisan camps mad at you simply means you can't do anything right.

Go ahead and put SpinSpotter on your toolbar if you want. (I might do the same; I'm actually kind of interested in what it identifies as "spun" news.) Just don't mistake it for anything that brings you any closer to understanding how news works or why some people do or don't trust it. If you have any questions, check out this fairly typical piece from the paid liars over at Fox and see if it doesn't qualify as "objective."

BusinessWeek would do well to stop buying magic beans, but that's a different matter.

6 Comments:

Anonymous Anonymous said...

As reported at somewhat greater length in the comments here, it appears that the SpinSpotter doesn't actually do anything at all, except perhaps to show you the passages that others have highlighted.

Or rather, it doesn't do anything at all, so far, except to bring in some money from credulous investors, who apparently hope to recover their stake from even more credulous investors down the road.

9:01 AM, September 09, 2008  
Blogger Strayhorn said...

I was wondering just the other day whether the NYT's revision on the direction of the sun was an example of lib'rul bias or corporate media spin.

And now I have an application to tell me. This modurn whirrled . . .

9:53 AM, September 09, 2008  
Blogger Cathy C. said...

I agree entirely. If these 66% don't trust journalists, why'll they be trusting a bunch of civilians who don't know what they're talking about. There's got to be a better solution than changing it for, pretty much, 2 or 3 people who may look at your changes. I feel like I'd just get irritated that my article looks like my 5th grade teacher bled her red pen on my gold star-worthy work.

7:57 PM, September 09, 2008  
Anonymous Anonymous said...

You have to understand this in the context of modern business fads. The in thing right now is "the wisdom of crowds". It's transparently obvious that ten thousand <s>ordinary people</s>unpaid volunteers with axes to grind will do a much better job than a professional editor would.

10:45 PM, September 09, 2008  
Anonymous Anonymous said...

The best thing to do is to use your mind, ask questions, and search for more information. That's more than any toolbar can do.

10:48 AM, September 10, 2008  
Anonymous Anonymous said...

good post

12:34 PM, September 21, 2008  

Post a Comment

<< Home